The Ubiquitin Ligase Triad1 Is Involved in Myelopoiesis and Interacts with the Transcriptional Repressors Gfi1 and Gfi1b.

We identified Triad1 as a gene that is upregulated by retinoic acid during the granulocytic differentiation of acute promyelocytic leukemia cells. In normal hematopoiesis, we show that Triad1 is weakly expressed in immature CD34+ bone marrow cells, and highly expressed in mature monocytes and granulocytes. Together, this suggests that Triad1 plays a role in the differentiation of hematopoietic cells. Triad1 contains a tripartite domain including two RING fingers, indicating that this protein might function as a ubiquitin E3 ligase, catalyzing the the conjugation of ubiquitin to substrate proteins thereby marking them for targeted degradation by the 26S proteasome.

Using GST pull down experiments, we show that Triad1 binds to the ubiquitin conjugating (E2) enzymes UbcH6 and 7. In addition, immunoprecipitation of Triad1 in cells that were transfected with FLAG-tagged ubiquitin shows that Triad1 binds to ubiquitinated proteins, and that Triad1 is capable of self-ubiquitination, further corroborating the assumption that Triad1 acts as a E3 ubiquitin ligating enzyme.

To study the role of Triad1 in hematopoiesis we overexpressed the gene in primary murine bone marrow cells using a retroviral vector that contains Triad1 in front of an IRES-GFP cassette. GFP positive cells were FACS sorted and used in colony assays (CFU-GM). Compared to empty vector controls (GFP alone), Triad1 expression resulted in more than 80% inhibition of clonogenic growth. Importantly, addition of the proteasome inhibitor MG132 (10E-8 M) reversed the Triad1-induced suppression of colony formation. Furthermore, three Triad1 expression constructs in which one of the conserved cys/his residues of the TRIAD domain (essential for function) were mutated did not show the suppressive effect on colony formation. Together, these data show that Triad1 is involved in myelopoiesis and acts through the ubiquitination of specific substrate proteins.

To identify these substrates, a yeast-two-hybrid screen of a human bone marrow cDNA library was performed using the Triad1 protein as a bait. Interestingly, the transcriptional repressor Gfi1b was found to bind to Triad1. The interaction was confirmed by immunoprecipitation using GFP-Triad1 and FLAG-tagged Gfi1b transfections in mammalian cells. We show that Triad1 binds to the zinc finger region of Gfi1b. This region is very (>98%) homologous to the paralogue Gfi1. Further immunoprecipitation analyses showed that Triad1 also binds to the zinc finger region of Gfi1. Gfi1 plays an essential role in neutrophil development and Gfi1 pointmutations result in neutropenia in man. Currently, we are studying the direct ubiquitination of Gfi and Gfi1b by Triad1 in in vitro ubiquitination assays. In addition, we are studying the effect of Triad1 on the transcriptional repression of the ELA2 and other promoters by Gfi1.